Re: IBM 5110 - Where does the character set live? And other questions.

[Robert via cctalk]

On Wed, Jul 12, 2017 at 03:13 AM, Christian Corti
 wrote:
> I hope this is also true in your case. According to your picture, the last
> column of each character is missing. It could be an issue around the shift
> register (e.g. the Display Data register, a latch, cold solder joint,
> ...). I just see that I have the following info on my site (didn't
> remember that detail ;-) ):

Apologies for the delayed reply and for any messy formatting. I seem
to have stopped getting messages from the list, so am building this by
replying to an earlier message and copying and pasting from the cctalk
archive.

It does look like that last column (or possibly two) is missing,
doesn't it? Going back to the info on your excellent site, the shift
register does seem a good candidate. If we assume that it is parallel
in, serial out, then the problem would have to be on the input side,
as all the bits pass through every flip flop (ruling them out) and the
output is flawed absolutely consistently (ruling out an intermittent
fault there). The same reasons militate against a serial input, I
think.

When I next see the board (hopefully this weekend) I'll try to
identify the shift register (which the cross ref on your site will
make a lot easier) and look for dry joints under high power
magnification. I don't have a complete and working logic analyzer and
in any case, the board is not really accessible with a logic probe but
TTL is cheap so, even if I don't see anything, a blind faith
replacement might be worth considering. I note that it's a 10 bit job
and there's nothing in your parts list greater than 8 bits, so I may
be looking at more than one IC.

Side note: It's probably not a good time to try out my shiny new heat
gun that I've never yet used. Maybe save my first go on it for
something more replaceable.

Of course, a parallel input implies a parallel source, so we still
can't rule out an issue with the character ROS. But I do feel like
we're making progress and I now have something to do that's more
focused than just pulling and reinserting boards until the Gold wears
off the pins, so I'm going to call that a step in the right direction.

Also, thanks for the additional scans, which I have already downloaded.

-- 
Robert

RE: HP 12653A line printer interface

[David Collins via cctalk]

I've looked for this card in the HP Computer Museum collection - no luck so
far. 

-Original Message-
From: cctech [mailto:cctech-boun...@classiccmp.org] On Behalf Of J. David
Bryan via cctech
Sent: Wednesday, 12 July 2017 4:12 AM
To: CuriousMarc 
Cc: Classic Computing List 
Subject: Re: HP 12653A line printer interface

Marc,


On Monday, July 10, 2017 at 22:24, CuriousMarc wrote:

> I thought I did, but what I have is the HP 12845B Line Printer 
> interface card, for which I could find the documentation.

Thanks for checking.  Yes, that does seem to be the more common card.  As
far as I know, the 12653A was used only for the HP 2767 (a rebranded Data
Products 2310), whereas the 12845B was used for a number of other HP
printers.


> Reading some more, it is meant for the 2607/261x series of printers, 
> which apparently use a narrower 7 bit interface (the 12566 is a 16 bit 
> interface card).

Which is all a bit odd, as the 2767 also uses 7 bits for data.  Unlike the
other printers that use differential interfaces, the 2767 uses single-ended
TTL-level (more or less) drivers and receivers, which may explain the use of
the microcircuit-based interface.

The 2767 signal drivers are adjustable for a 3- to 8-volt output level, so
perhaps the 12635A "modification" was to clip the inputs to avoid damaging
the standard microcircuit receivers (7400 TTL with an absolute maximum input
spec of 5.5 V).  In the absence of a manual, I was hoping that a photograph
would reveal the modification.


> But maybe you can inspire yourself from it.

The existing 2767/12653 simulation was reverse-engineered from the
diagnostic and OS drivers.  Although it works, I was hoping for something
more authoritative so that the code could serve as a reference for the
now-extinct hardware.

  -- Dave

Re: IBM 5110 - Where does the character set live? And other questions.

[Christian Corti via cctalk]

On Wed, 12 Jul 2017, Christian Corti wrote:
The scans I made are quite old; I do have a much better scanner now so I 
could just rescan the manuals for better quality.


Ok, I've added the 5114 MIM, and also added some pages of the System Logic 
Manual, including the Display Adapter and the 5114 drive. I will scan the 
other pages but it can take some time because the pages are very large and 
I need to scan them in slices.


Christian

Re: IBM 5110 - Where does the character set live? And other questions.

[Christian Corti via cctalk]

On Tue, 11 Jul 2017, Robert wrote:

So, faulty support logic, rather than a faulty ROS. That's encouraging.


I hope this is also true in your case. According to your picture, the last 
column of each character is missing. It could be an issue around the shift 
register (e.g. the Display Data register, a latch, cold solder joint, 
...). I just see that I have the following info on my site (didn't 
remember that detail ;-) ):
"The 8 bit wide characters are stored as a pattern of seven different bits 
followed by a 0-bit in a 2048x16 display ROS (character generator). These 
8 bits are loaded along with two 0 padding bits into a 10 bit shift 
register. The address of the 8 bit pattern is based on the Display Data 
register that contains the current character and the character row counter 
that counts the current display row within a character."


Christian

Re: IBM 5110 - Where does the character set live? And other questions.

[Christian Corti via cctalk]

On Tue, 11 Jul 2017, Santo Nucifora wrote:

I might have some better documentation that I just haven't had a chance to
scan yet.

[...]

5110 System Library Binder 1
SY31-0550-2 IBM 5110 Computer Maintenance Information Manual
SY31-0551-0 IBM 5114 Diskette Unit Maintenance Information Manual
SY31-0414-3 IBM 5103 Printer Maintenance Information Manual
SY31-0581-0 IBM 5110 Language Support Maintenance Information Manual
S131-0627-1 IBM 5110 Computer Parts Catalog
S131-0626-0 IBM 5114 Diskette Unit Parts Catalog
S131-0598-3 IBM 5103 Printer Parts Catalog
SY31-0553-1 IBM 5110 Maintenance Analysis Procedures

5110 System Library Binder 2
GA21-9300-0 IBM 5110 General Information and Physical Planning Manual
SA21-9311-0 IBM 5110 Customer Support Functions Reference Manual
SA21-9308-1 IBM 5110 BASIC Reference Manual

5110 System Library Binder 3
SA21-9306-0 IBM 5110 BASIC Introduction
SA21-9307-1 IBM 5110 BASIC User's Guide
SA21-9318-0 IBM 5110 Computing System Setup Procedure


This is more or less the complete manual set that I have, too. I also have 
the manuals for the Async/Serial I/O adapter and *maybe* the IEEE adapter.


The scans I made are quite old; I do have a much better scanner now so I 
could just rescan the manuals for better quality.


Christian

RE: IBM Scientific Subroutine Package (was Re: SIMH .tap file 7 track?)

[Dave Wade via cctalk]

Randy,
 To be pedantic, it never was ASCII it was EBCDIC. I would use the VTAPUTILS
from:-

https://sourceforge.net/projects/vtapeutils/

to convert it to".sws" format and then  read it on Hercules in an OS of
choice.

Dave

> -Original Message-
> From: cctalk [mailto:cctalk-boun...@classiccmp.org] On Behalf Of Randy
> Dawson via cctalk
> Sent: 11 July 2017 23:31
> To: Chuck Guzis ; General Discussion: On-Topic and Off-
> Topic Posts 
> Subject: IBM Scientific Subroutine Package (was Re: SIMH .tap file 7
track?)
> 
> Hi Chuck,
> 
> 
> Maybe you can answer a related question to the conversion of IBM 360 .tap
> files.
> 
> 
> I see the IBM FORTRAN Scientific Subroutine Package on Bitsavers in .tap
> format:
> 
> 
> http://bitsavers.trailing-edge.com/bits/IBM/360/360A_CM-
> 03X_Scientific_Subr_Pkg.zip
> 
> 
> How do I read/convert this back into the ASCII files?
> 
> 
> Thanks,
> 
> 
> Randy
> 
> 
> 
> From: cctalk  on behalf of Chuck Guzis via
> cctalk 
> Sent: Tuesday, July 11, 2017 3:10 PM
> To: CCtalk
> Subject: SIMH .tap file 7 track?
> 
> This is one that I haven't seen addressed.
> 
> If I'm reading a 7-track tape and writing a SIMH .tap file, what's the
custom?
> 6 bits per 8-bit byte, right-justified (i.e. 2 high bits zero-filled)?
> 
> Just wondering if there's a convention established for this.
> 
> --Chuck